Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

School Change: A Craft-Derived And Research-Based Strategy

by Gary A. Griffin & Susan Barnes - 1984

An intervention program, Changing Teacher Practice, was developed to cause effective staff developer behavior, effective teaching, and positive pupil outcomes. This study integrated knowledge about effective teaching and strategies for changing teachers and schools. (Source: ERIC).

The work reported herein was performed pursuant to a grant from the National Institute of Education, Department of Education. However, the opinions expressed do not necessarily reflect the position or policy of the National Institute of Education, and no official endorsement by the National Institute of Education should be inferred.

It is widely believed that research findings are not used as guides to school and classroom practice, Many reasons have been advanced for this condition, such as the reluctance of researchers to communicate their findings to constituent groups other than their research colleagues, the rigidity of school persons in terms of reconceptualizing the role of teacher or the organization of schools, the “closed-system” nature of many school settings, the lack of rewards for changing behavior, and the apparent lack of logical connections between craft and research findings.1 In short, it appears that the link between the research findings and the potential users of those findings has not developed to any sufficient degree.

The study reported here was an attempt to develop such a link. The following pages contain a description of an effort to alter staff-developer, teacher, and pupil behaviors.2


This study used an intentional intervention designed for persons responsible for in-service training of teachers. The intervention, termed Changing Teacher Practice (CTP), was based on the belief that the past two decades have seen an increase in the research-based knowledge about (1) effective teaching and (2) effective strategies for changing teachers and changing schools. The CTP study integrated these two bodies of research findings in an intervention strategy aimed at causing effective staff-developer behavior, effective teaching, and, ultimately, positive pupil outcomes.3

Briefly, the preliminary ideas for the CTP study might be put as follows: The findings from research on teaching have not found their way into classrooms. Likewise, the knowledge about how best to work with teachers and other school persons toward positive change has not been widely observed in practice. The research team considered these two bodies of research-derived information in terms of their potential for clinical in-service teacher education. The CTP study (1) used research on teaching findings as content for a planned intervention, (2) used research on teacher and school change as the basis for a delivery system for that content, (3) combined these two bodies of information into a specific change strategy, which was (4) introduced to staff-development persons in an ongoing school setting and (5) reinforced twice in that setting. Participants were observed for treatment effects on staff developer behavior, teacher behavior, and on-task student behavior.4


During the past years of intense and persistent attempts to introduce change into the instructional programs of schools, it was discovered that after an initial flurry of activity, observations of practice illustrated Sarason’s conclusion that “the more things change, the more they remain the same."5 With some variablity in terms of effect, it can be noted that these past attempts at change were also less than optimal in their impact on teachers and teaching. In the case of changing teacher behavior, it was suggested that the match between and among teaching practice, content differences, and the general nature of schools was such that change in a desired direction was unlikely.

An alternate explanation to those noted above is a simple but powerful one. It can be hypothesized that the attempts to change did not take into account two critical factors: teachers’ desire and/or need to change and the existential phenomena of schools, which must be manipulated in order for change to be seen as necessary and desirable from the perspective of teachers. Many strategies for school and teacher change advanced during the past two decades did not take into account the critical school variables that appear to be directly related to the degree of success of innovations. The CTP intervention employs a means by which school, system, and classroom variables are manipulated in such a way that change can take place and, ultimately, teachers can make the difference claimed by Good.6


A set of research-based conclusions about school change guided this effort. These conclusions, more often than not, have emerged as a consequence of post hoc analyses of efforts to innovate in schools. Some have come about as a conscious attempt to study change as it occurs. What is relatively rare is a prescription for change that is carefully documented and systematically subjected to rigorous research in contrast to prescriptions for change that rest primarily at the level of proposition or sets of “ought” statements. A cautionary word is necessary, however. One principal finding of change oriented research is that attempts to change are often situation-specific.7 That is, change efforts take place in contexts that vary widely. This contextual variation apparently produces comparably unpredictable results unless the change strategy is structured to account for such a contextual variation. Therefore, the CTP intervention had as a deliberate component the opportunity for staff developers to select strategies and teaching behaviors that, after systematic reflection about situation-specific school variables, fitted their settings.

The intervention was implemented with the staff-development persons in such a way that the conclusions from the change literature were consciously attended to, while at the same time the persons charged with system staff development made the primary, situation-specific decisions about how to move forward with the change strategy.

The second point is important. In contrast to some school-improvement efforts, the CTP study did not prescribe what instructional leaders must do with their teachers. Instead, the strategy proposed a set of options for working with teachers and focused participants’ attention on the need to make reasoned and rational selections from the set. Likewise, staff developers were not required to focus on any or all of the research-derived teaching behaviors but to select the ones that were most appropriate for the teachers with whom they worked.

The point of intervention for the CTP study was a critical issue. Because the Research in Teacher Education (RITE) research team was concerned about the replicability of the CTP study in settings other than the one in which the original research took place, it was believed that the most appropriate intervention point was at the staff-developer level. That is, the content and the strategy would be introduced to persons charged with staff development who would be studied in terms of their own behaviors and beliefs as well as their apparent effects on the teachers with whom they worked. The CTP study was based on the assumption that school leaders can provide instructional leadership to faculty groups if they are helped to focus their behavior and if they have a knowledge base from which they can work.


Certain specific classroom teaching behaviors have been identified by research as being positively related to higher student scores on standardized mathematics and reading tests. Research has also indicated that student achievement is related to on-task behavior of students. These findings formed the content of the in-service training to be delivered by staff developers to the teachers. In addition, certain specific features have been identified by research studies as positively related to the facilitation of desired change in teacher practice. These features informed the system for delivery of the content. Taken together, these two bodies of knowledge formed synergetic in-service training that was research-derived, theoretically sound, conceptually coherent, and directly related to the craft of doing schooling.


The teacher behaviors forming the content of the in-service training have been identified by the line of research generally referred to as “teacher-effectiveness research."8 In these studies, teachers and students in a variety of settings were observed for periods of time and their behaviors recorded. From these observations of teaching craft, teacher behaviors that related to student learning as measured by standardized tests were then identified. An operational definition of the effective teacher emerged from this line of research: The effective teacher is the teacher whose classes regularly score higher on standardized achievement tests than do classes of other teachers after entering differences among classes are statistically controlled.9

Descriptions of Selected Studies

Three lines of research provide most of the findings selected for inclusion in the CTP intervention. The major topics of interest in one group of studies were management and organization of the classroom. The work of Brophy and Evertson on the Correlates of Effective Teaching Project began with correlational studies in the second and third grades.10 Other studies based on their work eventually moved into other grades and subjects, including an experimental study in reading at the first-grade level and an experimental study in mathematics and English at the junior high level.

The second set of studies, which began with the work of Good and Grouws11 on the Missouri Mathematics Effectiveness Project, were correlational studies of teachers and students in third- and fourth-grade mathematics. Later, experimental studies based on these findings were conducted in sixth-, eighth and ninth-grade mathematics. Major emphasis in these studies has been on systematic instruction.

Another line of research began with the work of Stallings on the Follow Through evaluation of the Planned Variation programs.12 Data for the correlational study were collected for reading and mathematics in grades one and three. These findings served as a basis for extending research into the teaching of basic reading skills in grades seven to twelve. Both a correlational study and an experimental study were conducted at this level. The emphasis in these studies has been on a program of effective instruction in reading skills.

Other classroom studies contributed in some way to the CTP intervention. The work on the Follow Through evaluation in Florida by Soar and Soar13 and the work by McDonald and Elias14 on the Beginning Teacher Evaluation Study addressed the importance of academic time.

This research on teaching from large-scale, classroom-based studies provides a profile of only one definition of an effective teacher, where effectiveness is measured by student outcomes on standardized tests. In addition, the profile is limited to areas of teaching that have recently received heavy research emphases: learning environment, management of behavior, classroom instruction, and teaching style. While it is recognized that these areas do not form a complete picture of teaching, the findings indicate that some teaching behaviors are associated with increased student achievement in mathematics and reading at the elementary-school level. The effective teachers in these studies tended to establish a work orientation in the classroom while maintaining a warm, supportive environment. They also were well organized and emphasized management of the classroom in order to optimize the productive use of time. During class, effective teachers stayed actively involved with students to prevent misbehavior and intervened promptly to stop misbehavior. When presenting new material, effective teachers used a systematic instruction plan that included gaining students’ attention before beginning the lesson, making a clear presentation, allowing students to practice new skills, monitoring work and providing feedback, assigning individual seatwork, and evaluating students’ responses. Effective teachers generally interacted with the whole class during class time and moved students through discussions at a brisk pace with a high level of student success.

These research findings support many common teaching practices and answer criticisms that educational research has no relationship to real classroom problems and situations. While there are other conceptions of good teaching that view teaching as an art or as a job requiring certain philosophical or psychological orientations, the CTP study was concerned only with recent findings in classroom-based research and not the assumptions underlying this particular approach to the study of teaching. At present, if school emphasis is on achievement in basic skills as measured by standardized achievement tests, this research provides teachers with behaviors that facilitate increases in achievement and still allow adaptation to fit particular classroom needs.


Given the stated teaching behaviors as the desired outcomes of a program of in-service training for teachers, the next logical step was to facilitate change in teacher practice in the direction of greater incidence of the stated behaviors. There are, to be sure, problems inherent in the formulation and testing of a strategy to alter teachers’ pedagogical practices. Good, among others, acknowledges the context-effect issue. 15Ward and Tikunoff identify problems related to “working on” rather than “working with” teachers.16 Griffin characterizes some of the issues of evaluating teacher-change efforts.17

Three widely and frequently cited research efforts provided extensive data regarding changing teacher practice: (1) the I/D/E/A five-year study of change in individual schools;18 (2) the Rand Corporation study of federal programs supporting educational change;19 and (3) the Concerns Based Adoption Model (CBAM) work on the implementation of innovations.20 Particular emphasis was placed upon Sarason’s notion of institutional regularities,21 Goodlad and Klein’s hypotheses regarding why change is blunted in schools,22 Bentzen’s findings derived from the I/D/E/A study of school change,23 and Berman and McLaughlin’s propositions about system characteristics that foster and support efforts to change.24 In addition, and related to the I/D/E/A findings and processes, potential strategies were derived from organization development and practice.

Research on the process of change provides a description of various practices found to be positively associated with the successful implementation of innovation. The descriptive listing is neither exhaustive nor absolute but is logically generalizable and predictable in terms of expected results.

From studies of school and teacher change it can be inferred that effective leadership will take into account:

1. opportunities for teacher interaction focused on professional issues

2. provision of technical assistance to teachers

3. adaptation of ideas and programs toward a “fit” with school and classroom regularities

4. opportunities for reflection

5. focused and precise (rather than general) attention on important school issues

These broad areas of leader behavior were more sharply defined in operational terms and presented as part of the CTP intervention in the form of a list of desired staff-developer behaviors.

Methods and Procedures

As already noted, the CTP study consisted of an intervention with staff developers to provide them with strategies and content to improve reading and mathematics instruction by elementary-school teachers. To assess the impact of this intervention, a treatment group-control group quasi-experimental design was used. The research question toward which the methods and procedures were directed was: What differences in behaviors of staff developers, teachers, and students can be attributed to the CTP intervention?


Because the CTP study was part of the research agenda of a national research and development center, it was considered important that it reflect, as much as possible, the realities of the nation’s schools. In that a major national educational issue is the multicultural nature of the school population, the CTP study was implemented in a school system reflecting multicultural student and educator populations.

School improvement is not the sole responsibility of any one agency in the educational community. Teachers, administrators, higher education persons, and researchers come together to better understand the ways and means of schooling. Therefore, the CTP study was conducted in a school setting where there was an active set of professional associations (teacher organizations, administrator organizations, etc.) and some link between the school system and other institutions with similar purposes (e.g., colleges and universities).

The school system had also a demonstrated history of attending to the issue of school improvement. Although we believed the study described here had promise to bring about changes at several levels of the school organization, the staff was aware of the necessity that any change effort be placed in what might be called a “receptive” environment. This meant that the school policies and institutional climate would be amenable to change and to reflection.

The persons who agreed to participate in the study, especially staff developers, had some flexibility in terms of the use of their time and in terms of their assignments. This flexibility was part of the agreement between the research team and the officers of the participating school system.

It was also desirable that the staff-development activities in the participating system take place at the individual building level. This feature corresponded to several of the tenets of school change strategies and, as well, provided the research team with a more economical way to accomplish the logistics of data collection.

Finally, and most importantly, system personnel at staff-developer and teacher levels agreed to participate in the study. This agreement was supported by policy-level persons in the school system and in the professional associations.

A representative group of ten individuals responsible for working with teachers in an effort to promote more effective teaching and school improvement in the selected district were identified. Each was assigned to one group of five, matched as closely as possible according to role, socioeconomic status (SES) of school, prior experience, years in the position, and reputation for effectiveness. One of each pair was randomly assigned to the treatment or control condition at the staff-developer level.

Two teachers from among the group with whom each staff developer worked were selected by staff developers. Those ten teachers working with the staff-developer treatment group constituted the teacher treatment group, the other ten constituted the teacher control group. Students of the treatment group teachers constituted the treatment-group students. Students of control group teachers constituted the control-group students.

Similarities within and across the various role groups participating in the study appeared to outnumber observed differences. The staff developers in both treatment and control groups were over thirty years of age: all but one had a Master’s degree (and that exception was nearing completion of a Master’s). Each group of staff developers consisted of three resource teachers and two school principals; all but one in each group were school-based. All staff developers had prior classroom teaching experience at the elementary level. The treatment- and control-group teachers showed similarities as well. For both groups, most were Anglo women, and at least half the teachers in each group had Master’s degrees. Roughly half the teachers in each group were in primary classrooms, and about half in each group belonged to professional associations. Seven teachers in each group reported working in medium-sized schools. Also, the teachers responded in like manner (no statistically significant differences) to the open-ended questionnaires about their teaching abilities and the abilities of their students.


Because the CTP study was concerned with the effects of the staff-developer behavior on teachers with whom they worked and with the effects of the teacher behavior on pupil outcomes, the independent-dependent variable relationship was conceptualized as shown in Figure 1.

The RITE staff expected change in staff-developer behavior (dependent variable) and, further, as a consequence of staff-developer behavior (now considered an independent variable) teachers were expected to behave in certain ways (dependent variable). Likewise, we expected teacher behavior (now an independent variable) to affect pupil outcomes (dependent variable).


The intervention provided to treatment-group staff developers took place three weeks prior to the opening of the 1981/1982 school year. For five consecutive days the treatment-group staff developers met with the research team in a hotel conference room near the research sites. The first day of the intervention was devoted to an explanation of the study and to a set of opportunities to learn about research-based effective teaching, using print materials, videotape, discussion, and narratives of classroom activity. Participants were also given a collection of print material that ranged, from research reports through syntheses of research study findings to manuals designed for teacher use in planning and organizing instruction. The second day of the intervention was planned to acquaint the participants with research-derived effective leadership behavior. Three general areas related to desired staff-development behaviors were addressed: mutual adaptation, organizational development, and support networks. For planning, learning, observing, and descriptive purposes, a list of specific staff-developer behaviors, derived from the research literature related to teacher change, was provided to the staff developers. The third day was devoted to learning about the use of the Concerns-Based Adoption Model in order to help participants (1) understand the process of change, (2) correctly diagnose teachers’ concerns about change, and (3) plan for effective interventions to bring about desired change. The fourth and fifth days of training were given to planning for site implementation. Participants answered in writing a set of questions that revealed information about their individual school sites. The answers were then used to design a plan to work with teachers toward increased frequency of research-based effective teaching behaviors chosen for emphasis by the treatment staff developers.



All staff developers and teachers, treatment and control group, were asked to keep a written account of each staff-development interaction. They were asked to include, as a minimum, information about the type of interaction, the content, and the outcomes.

Teachers were observed using a low-inference observation instrument, the Barnes Teacher Observation Instrument. The observations, beginning on the first day of school and ending in February, focused on several categories of teaching behavior: planning and preparation, presenting content, interactions, conducting practice, conducting seatwork, holding students responsible for assignments, organizing the classroom, presenting rules and procedures, holding students responsible for behavior, and reacting to student behavior. The record of observed teacher behaviors documents the frequency of desired behaviors, as well as the naturally occurring sequences of these behaviors.

To provide a “snapshot” of classroom activity during the observations, observers recorded information at ten-minute intervals on the student engagement rating form. That information included (1) the format of the class, (2) the number of adults in the classroom, (3) the subject matter, (4) the total number of pupils in the room, (5) the number of students engaged in academics, procedures, or off-task behavior, and (6) a rating of apparent student success.

In addition, several questionnaires were administered to staff developers and teachers in order to gather information about their backgrounds, their current professional situations, their perceptions related to their own abilities, and their thoughts regarding student ability to learn.



Treatment-group staff developers demonstrated slightly more than twice as many of the desired staff-developer behaviors as did the control group. The results of a Mann-Whitney test indicate that this difference is statistically significant at the .01 level. The within-group means were also in favor of the treatment group in that the highest mean for the control group was lower than the lowest mean for the treatment group.

The frequencies of each of the desired staff-developer behaviors provide the substance of the difference between the group means. The strongest differences in favor of the treatment group (there were no instances of frequencies that favored the control group) were for diagnosing school- and classroom-specific regularities; providing teachers with opportunities to interact with one another about teaching and schooling; providing teachers with opportunities to plan together; providing teachers with feedback that is objective, concrete, and focused; adapting staff-development behavior according to personal and organizational characteristics of “users”; demonstrating knowledge of effective teaching as revealed by research; working with teachers on adaptation of teaching strategies; linking teachers to technical assistance outside the immediate school; including the building principal or resource teacher in teacher in-service activities; reflecting on effects of one’s own behavior; and focusing on teacher behavior.

That these behaviors account largely for the difference between the treatment- and control-group staff developers is of interest because they are, in great measure, characterized by (1) attention to the school as an organization, (2) reconsideration of the so-called isolation of teachers from other teachers and administrators, (3) recognition of the need to adapt behavior and expectations according to observed school conditions, and (4) promotion of interactions within and outside the school environment. Taken together, the behaviors are among those that break down the perceived inadequacies of schools and school people to act together toward organizational goals.


In the categories dealing with academics, the treatment teachers exhibited higher mean rates in “Planning and Preparation,” “Presentation,” and “Holding Students Responsible for Assignments.” Planning and preparation for academics was seen twice as often in the classrooms of treatment teachers as in the classrooms of control teachers (p < .09). For academic presentation, teachers in the treatment group again exhibited a statistically higher mean (p < .01). Although treatment-group teachers had higher mean rates for holding students responsible for assignments, the differences were not statistically significant. Control-group teachers produced higher mean rates for “Interactions, ” “Practice,” and “Seatwork”; however, none of the differences were statistically significant.

A more consistent pattern was seen in the four observation categories related to classroom management. In each of these categories, the treatment-group teachers displayed the higher mean rates. For two categories, “Holds Students Responsible for Behavior” and “Reactions to Students Behavior,” the differences were slight. For the remaining categories, “Organizes Classroom” and “Presentation of Rules and Procedures,” the differences were statistically significant (p < .03 and p < .06, respectively).

In some observation categories— “Planning and Preparation for Academics” (p < .06), “Practice” (p < .04), and “Seatwork” (p < .04)—significantly higher or lower rates of teacher behaviors were associated with particular staff developers. In other words, individual staff developers had significant effects on the pairs of teachers with whom they worked. In some categories— “Interactions” (p < .003), “Seatwork” (p < .08), and “Holding Students Responsible for Assignments” (p < .08) —the mean rates of teaching behaviors increased significantly over time (p < .001) for both treatment and control groups. Mean rates for “Presenting Rules and Procedures” decreased significantly over time, as would be expected. Finally, significant differences in content were usually associated with higher rates for mathematics (“Practice,” “ Seatwork,” and “Holding Students Responsible for Behavior”). The only category for which reading produced significant content effects was the “Interactions” category.


It was noted in the discussion of staff-developer effects that treatment-group staff developers reported using significantly more of the desired staff development behaviors than did control-group members. This difference was accounted for largely by a subset of eleven of the twenty-three behaviors. In the discussion of effective teaching behaviors it was reported that treatment group teachers demonstrated significantly more of the research-derived behaviors in certain categories than did control-group teachers. Other categories of teaching behaviors were not significantly different in favor of either the treatment or control group.

Examination of these findings suggests that there is a relationship between effective leadership behavior and teaching. This relationship is not clear-cut and is not readily understood, given the methods and procedures used in this study. What seems to have happened with the treatment-group staff developers is that they used the relatively unconventional leadership behaviors (e.g., adaptation of their own behaviors according to their teachers, provision of technical assistance outside the school, providing teachers opportunities for interacting and planning together, etc.) and focused those behaviors on teaching. The finding that there was such a dramatic difference between treatment and control staff developers in terms of focusing on teaching supports the possibility that the CTP study provided school leaders with a body of knowledge about classroom life that was coherent, organized, and, one suspects, meaningful to teachers.


Higher percentages of treatment-group students were on-task in academic activities, and higher percentages of control-groups students were on-task in procedural activities. In neither case were these differences statistically significant. Both treatment- and control-group students did exhibit significant increases in their percentage of on-task academic behavior over time (p < .03) and significant decreases in their percentage of on-task procedural behaviors over time (p < .04). However, because of the non significant differences between the percentages of treatment- and control-group students engaged in on-task behaviors, the CTP intervention appeared not to have a marked effect on pupils.


There are several ways that these findings can be understood, if not explained fully. Among them are perspectives about the setting and the participants, the conventions of schooling, and the conventions of teaching.


Earlier it was noted that the research team deliberately selected a school district that was characterized by complexity, reflective of problems and issues faced by large numbers of other districts, and with a history of concerted effort in staff development and school improvement.

The district is large, covering the spectrum of SES characteristics, and has a recent history of making vigorous attempts to respond to the problems that arise from providing equal educational opportunity to a heterogeneous population. Some schools in the district are new and architecturally pleasing; others are old and in various states of disrepair. There is a massive busing program and a wide variety of special programs of both organizational and academic natures.

Of particular importance to the CTP study was the district’s recent several year history of attempts to increase student achievement as measured by standarized achievement tests. Among the strategies used to accomplish this objective were systematic attention to increasing student time-on-task, providing teachers with information about how to decrease classroom distractions, and the introduction of a carefully designed curriculum in reading and mathematics. These districtwide efforts were important because, in effect, they were focused on many of the same dimensions of schooling as was the CTP strategy. That is, they zeroed in on the importance of pupil instructional time and the relation of teacher behavior to that time. This was particularly true of the curriculum packages that used much of the teaching research as bases for what students and teachers were directed to do during reading and mathematics instruction. The teachers who used the curricula were allowed little flexibility in that decisions about lesson plans, materials of instruction, testing, and grouping were made for them.

Consequently, in more than half of the schools and classrooms in the CTP study, equally divided between treatment and control groups, there was already in place a massive effort to increase student achievement through the use of research findings and curricula based, in large measure, on those findings. In research terms, it might be said that the district program was Treatment A and the CTP study was Treatment B.

It is possible that some of the findings of the CTP study can be understood in terms of these setting characteristics. The partial effects at the teacher level may be an artifact of the concerted effort that had already made an impact on participating teachers. If this is the case, the effects that were achieved through the CTP intervention can be considered even more positively.

The significant effects at the staff-developer level are also partially understood as a consequence of knowing about the setting. Although there had been concerted effort aimed at teachers, there had been less systematic attention given to research-derived leadership strategies. This was particularly true of principals, who participated in their own in-service activities; these activities were, however, according to participant reports, more organizational and procedural than substantive. It was less true of resource teachers in the study in that these persons were participants in the development and use of the curricula noted earlier. The resource teachers were experts in the curricula but were not necessarily given help in the most effective ways to guide teachers in their use of those curricula.

It was noted that the participants in the study, staff developers and teachers in both treatment and control groups, were, for the most part, experienced, well-educated, and highly regarded professionals. The district takes pride in the caliber of its staff and has a history of rewarding excellence. It is possible that the participants, much like the setting, are less typical of conventional school districts than was originally suspected.

The nature of the setting and the characteristics of the participants, then, worked for and against testing the power of the CTP strategy. On the one hand, there were setting variables (e.g., the history of attention to teaching research) and participant characteristics (e.g., the number of advanced degrees held by teachers) that may have mitigated against strong effects. On the other hand, the lack of systematic district attention to what research can tell the instructional leader or the notions inherent in “teaching” classroom rules and procedures probably increased the potential of the strategy for illustrating differences between treatment- and control-group participants.


It is axiomatic that school people “take care of business.” The ways in which this is done have been described in some detail and the pictures that emerge from the descriptions are of more similarity than difference across school sites. The persistent rhetoric about how instruction can be influenced by school leaders is, to many observers, more hollow than not.

Research into effective schools and the school change process, however, has identified “outlier” school leaders, persons whose leadership behavior is identified with positive change, good climate, a sense of professionalism, and student accomplishment. Examination of the research and proposition regarding school leadership suggests the possibility that, to a certain extent, the “effective” instructional leader is someone who pushes back against the conventions of schooling, someone who reconceptualizes the organizational and management structure of schools and uses that reconceptualization as a means to make change happen.

The CTP intervention was most effective in terms of causing staff developers to act differently than might be expected if business as usual were carried forward. That is, the greatest differences between treatment- and control-group staff developers were largely linked to those leadership behaviors that are not conventional ones. Providing teachers opportunities to work together and plan together mitigates against the isolation of teachers from one another and may recreate a shared sense of mission. Understanding what makes schools and teachers different from other schools and other teachers may have suggested more powerful in-service activities. Consideration of how one’s own behavior, and the behavior of teachers, must be modified to adapt to the demands of schooling may have increased sensitivity to the problems of change. Recognizing the available and pertinent resources inside and outside of the school (human and technical) may have provided a broader picture of what can be accomplished in classrooms. And, reflecting on one’s own work in terms of effects, rather than only in terms of whether an activity was completed, may have provided stimuli for redirection of effort.

There were, however, certain research-based leadership strategies that were not accomplished as a consequence of the CTP intervention. These, too, can be understood in terms of the conventions of schooling. For the participants in this study, for example, provision of technical assistance in the classroom did not take place to any large degree. There is little in the histories of schools to suggest that this is either easy or, in fact, feasible in terms of the ways that schools are conventionally organized. Similarly, diagnosing and acting on teachers’ concerns is a powerful strategy that needs considerably more understanding and skill than the CTP strategy was able to accomplish. Further, the organization of schools and the dailiness of school activity probably mitigates against the concerted effort and time needed to use concerns theory systematically or thoroughly.

As was true in the discussion of the findings in relation to the study’s setting and participants, the conventions of schooling may account for the success of the strategy in that some of the staff-developer strategies were possible to effect because they were relatively closely aligned with doing business as usual. Others, however, may have been so dramatically unconventional that they required more system accommodation than the CTP intervention could induce.


It is sometimes forgotten that the research-based effective teaching behaviors were, for the most part, invented by teachers. That is, teachers engage in certain practices that researchers discover to be consistently related to some positive student outcome, most often achievement gain on standardized tests. The CTP study was an attempt to bring together these research-based effective teaching behaviors in such a way that staff developers could use them as the content of their work with teachers. Because the teaching behaviors are, in the main, neither exotic nor inconsistent with conventional classroom activity except in terms of focus or emphasis, it is not surprising that there were statistically significant effects for some groupings of behaviors and not for others.

The intervention used with the staff developers affected the teaching behaviors of the treatment-group teachers in the desired direction, in some cases significantly. The greatest differences in teaching effects were seen in two important components of the teaching process: planning and presenting. These differences were seen whether the content of that teaching process was academic or whether it was rules and procedures for classroom activity. The differences for planning are probably tied closely to one particular intervention resource, Organizing and Managing the Elementary School Classroom.25 Following suggestions made in the manual would logically result in the posted rules or academic schedules seen in classrooms of treatment-group teachers. Use of the manual would also account for the increased rates for presenting rules and procedures as the content of a “lesson,” an idea that may be foreign to many teachers.

The increased rates for academic presentation are more difficult to account for, and according to some research more difficult to accomplish. Two ideas should be considered. An intervention resource was the Teacher’s Manual: Missouri Mathematics Effectiveness Project,26 which stressed the development of the conceptual ideas of the lesson through teacher presentation. One could conclude that the manual achieved its purpose in the study by increasing the presentation behaviors of the treatment teachers. Good and Grouws, however, expressed some concern with their success in enhancing the development portion of mathematics lessons with the teachers in their own study.27 The key here is one of quality versus quantity. Good and Grouws were interested in the quality of teacher presentation skills (clarity, for example) while this study was concerned with the quantity or frequencies of certain teaching behaviors. It may have been easier to simply increase frequencies of behaviors in a teacher’s repertoire in this study than to increase the variety of “effective” teaching behaviors in the Missouri Mathematics Project.

The conventions of teaching also help us to understand other findings in the CTP study. Significant content differences, for example, appeared to be logically related to common practice (i.e., more practice, seatwork, and presentation in mathematics) or to be the result of observer coding decisions (i.e., coding round-robin reading as interactions related to product questions). Likewise, the persistence of seatwork as a classroom activity probably precluded any systematic difference between treatment- and control-group teachers. It is unlikely, then, that some of the research-based teaching behaviors would be new to the setting or to the participants. In light of the discussion of the school district’s attention to research on teaching in recent years, a statistically significant difference for all teaching behaviors in the study between treatment and control groups would be even more surprising.

Returning to the notion that teachers invented the effective behaviors and researchers found them to be effective according to a criterion, it is encouraging that certain groups of teaching behaviors were statistically significant in favor of the treatment group. There is little in the setting to suggest that influences other than the CTP intervention caused the differences.


There are several ways that the CTP study can be viewed in terms of implications for the improvement of school practice. The two that will be discussed here are implications for research into practice and local capacity building.


There has been considerable talk and some activity around the topic of the so called gap between research and practice. It has been assumed that the ideological and temporal distances between what research discovers and what practitioners do are almost insurmountable. The CTP study, a deliberate attempt to bridge the gap, resulted in a set of desired outcomes that suggest that practitioners will use the results of research. Several implications can be drawn from the implementation of the CTP intervention.

First, practitioners in this study were neither disdainful of nor resistant to lessons learned from systematic inquiry. One can speculate that the reasons for their positive responses to the CTP intervention were based largely on the fact that the research that was presented was directly related to practice and, in almost every instance, was “commonsensical” in terms of practitioners’ views of their world.

Second, the research provided a focus for doing the work of classrooms and schools. It provided conceptualizations of instructional leadership and teaching that were coherent and that could serve as rallying points around which to organize practice. Although there is an obvious public emphasis on somehow improving schools, there is a less obvious response from either the research or practitioner community regarding how to go about that improvement. The CTP intervention seemed to provide bases for improvement activity, a welcome resource for school persons.

Third, the research was translated from the sometimes painful jargon of the research community into the more conventional language of schools. This translation was accomplished in some instances in print by the resource materials provided to participants and in others interpersonally by the RITE research team during the intervention. Although the study did not specifically address this issue, participant self-reports at the conclusion of the study testified to its importance in their decisions to adopt the strategy in schools and classrooms.

Fourth, the CTP strategy was conceptually and practically linked to ongoing processes and expectations in the school setting. As discussed earlier, the school district had in place activities and statements of purpose that were ideologically and theoretically aligned with the intentions of the CTP intervention at system, school, and classroom levels. As with other speculations in this section, it can be surmised that this alignment was a significant system variable contributing to the overall success of the CTP strategy.

Fifth, the CTP strategy was not a set of prescriptions for action. That is, participants were not required to engage in certain practices. Rather, participants were expected to analyze their own situations, using a relatively rationalistic set of procedures and data sources, and then make decisions and act on their analyses. This is in marked contrast to improvement strategies that demand fidelity to a set of ideas or practices. The CTP strategy’s demand was to consider and act on the perceived match between school/classroom characteristics and a set of research-based options for leadership and teaching practice.

In sum, the implications of the implementation of the CTP strategy in terms of research into practice suggest that similar interventions will be positively viewed and acted on when research is seen as directly related to the problems and issues in the setting, is relatively familiar to participants in term of practical activity, is understandable conceptually and linguistically, and can be subjected to adaptation depending on the character of the setting.


The CTP study was, of course, an attempt to strengthen the role of research findings as guides to ongoing leadership and instructional practices. The design of the study was such that another purpose could be served, that of increasing local school district capacity to improve classroom activity. Although there are many entry points into schools for improvement purposes, the CTP strategy elected to make its primary impact at the school-leader level so that influence on the participating district could be deep and wide. This is in contrast to a strategy, for example, that would work with individuals or groups of teachers. The CTP strategy was constructed on the assumption that staff developers work with large numbers of teachers who, in turn, work with large numbers of students. If the strategy was effective, (1) it could reach more people than might be possible or feasible when working only with teachers and (2) the effects would increase the capacity of the district to do school improvement because of the enhanced knowledge and skills of its leadership cadre. As was noted earlier in this report, it is possible that the CTP intervention provided school leaders with a “technical core,” a body of knowledge and skill required for cohesive and successful organization. Further, the CTP strategy’s design for staff developer-teacher linkage acted on the phenomenon that has been called “loose coupling.“28 School systems have been criticized because their organizational components are not tightly joined to serve common missions. In the CTP strategy, two organizational components, leadership and teaching, were tightly coupled through systematic attention to what research has shown to be effective teaching.

It has been shown that the strategy was largely successful in accomplishing the goal of increasing the frequencies of certain leadership behaviors used by staff developers. The implications of this for practice are obvious. The behaviors are not specific to the content of the CTP intervention. They can be used in both maintenance and improvement activities. They can be used for a wide variety of purposes within the school organization. They are economical in terms of material resources. There is no magic involved. And, because they are carried by people rather than equipment or technology, they can be transferred from one place in the system to another.


The implementation of the CTP study demonstrated the possibilities (and the problems) associated with attempting to introduce research findings into school and classroom settings through an intervention aimed at school leaders. The study accomplished many of its goals and fell somewhat short of others. The reasons for both accomplishments and shortfalls are as complex as are the schools and classrooms that were the ultimate targets of the strategy. The intervention was a research success, although the success was not as dramatic as might have been desired. At the same time, it must be acknowledged that the ambitions of the study were large ones.

This report has centered, quite naturally, on the effects of the CTP intervention in research terms. The effects in more practice-related ways provide a somewhat different chronicle.

The participating school district officers were enthusiastic about the CTP strategy from the beginning. There was high level of excitement about the possibilities before, during, and after the intervention and subsequent data collection. Participant response from treatment-group members was consistently positive. Importantly, participants testified that other staff developers in the district should have the same opportunity they had been provided.

When the study was completed and the findings were available, the district adopted the CTP intervention for 175 elementary- and middle-school principals and resource teachers. Part of the adoption decision was made on the basis of research. That is, the findings of the CTP study were positive enough to suggest the usefulness of the strategy for district purposes. A large part of the decision, however, appears to have been made on more practical grounds. The RITE research team suspects that the commonsense approach to doing leadership and doing teaching was compelling to district officials. Also, the cost-effectiveness of the strategy, in terms of financial resources, was undoubtedly attractive and the probable generalizability of the leadership behaviors to school-improvement purposes other than teaching was a factor.

When the CTP research was completed, the research team returned to the study site and demonstrated the intervention, modified somewhat to fit current district needs and schedules, and local persons were selected to carry it forward. These persons had been members of the original treatment-group staff developers and were sensitive to, knowledgeable about, and skillful in the content and processes of the intervention. Since the final analyses of the data presented in this report, the intervention has been presented to more than 100 principals and resource teachers, in groups of 10 to 15, and the principals and resource teachers have worked with more than 3,000 teachers. This was accomplished in a five-month period.

The CTP study was a powerful demonstration of the appeal for school people of research knowledge when it is conceptually and practically related to doing the business of schools. It is too often forgotten that recent research findings regarding effective teaching and leadership are not products of the fertile minds of researchers. They are, instead, largely inventions of teachers and school leaders; the researchers discovered them and showed how they were related to student cognitive gain. This craft-to-research-to-improvement paradigm has strong potential for supporting necessary and important school and teacher change activity.

Cite This Article as: Teachers College Record Volume 86 Number 1, 1984, p. 103-123
https://www.tcrecord.org ID Number: 930, Date Accessed: 1/28/2022 3:29:17 AM

Purchase Reprint Rights for this article or review
Article Tools
Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • Gary Griffin
    University of Illinois, Chicago Circle

  • Susan Barnes
    University of Texas, Austin

Member Center
In Print
This Month's Issue